app.py

"""
References:
- https://medium.com/@turna.fardousi/building-a-multimodal-chatbot-with-gemini-api-8015bfbee538
"""

import os
import time
from typing import List, Tuple, Optional
import google.generativeai as genai
import gradio as gr
from PIL import Image
from dotenv import load_dotenv

load_dotenv()

GEMINI_API_KEY = os.getenv("GEMINI_API_KEY")
genai.configure(api_key=GEMINI_API_KEY)

TITLE = """<h1 align="center">🎮Chat with Gemini 1.5🔥 -- Beta Preview</h1>"""
NOTICE = """
Notices 📜:
- This app is still in development
- Some features may not work as expected
"""
ABOUT = """
Updates (2024-8-12): Created the App

Info:
- Model: Gemini 1.5 Flash
- Features:
    - Langchain integration
    - Google search
"""
ERRORS = """
Known errors ⚠️:
"""
FUTURE_IMPLEMENTATIONS = """
To be implemented 🚀:
- Select other Gemini / Gemma models
- Upload files
- More tools other than web search
"""
IMAGE_WIDTH = 512

def preprocess_stop_sequences(stop_sequences: str) -> Optional[List[str]]:
    return [seq.strip() for seq in stop_sequences.split(",")] if stop_sequences else None

def preprocess_image(image: Image.Image) -> Image.Image:
    image_height = int(image.height * IMAGE_WIDTH / image.width)
    return image.resize((IMAGE_WIDTH, image_height))

def user(text_prompt: str, chatbot: List[Tuple[str, str]]):
    return "", chatbot + [[text_prompt, None]]

def bot(
    google_key: str,
    image_prompt: Optional[Image.Image],
    temperature: float,
    max_output_tokens: int,
    stop_sequences: str,
    top_k: int,
    top_p: float,
    chatbot: List[Tuple[str, str]]
):
    google_key = google_key or GEMINI_API_KEY
    if not google_key:
        raise ValueError("GOOGLE_API_KEY is not set. Please set it up.")

    text_prompt = chatbot[-1][0]
    genai.configure(api_key=google_key)
    generation_config = genai.types.GenerationConfig(
        temperature=temperature,
        max_output_tokens=max_output_tokens,
        stop_sequences=preprocess_stop_sequences(stop_sequences),
        top_k=top_k,
        top_p=top_p,
    )

    model_name = "gemini-1.5-flash" # if image_prompt is None else "gemini-pro-vision"
    model = genai.GenerativeModel(model_name)
    inputs = [text_prompt] if image_prompt is None else [text_prompt, preprocess_image(image_prompt)]
    
    response = model.generate_content(inputs, stream=True, generation_config=generation_config)
    response.resolve()

    chatbot[-1][1] = ""
    for chunk in response:
        for i in range(0, len(chunk.text), 10):
            chatbot[-1][1] += chunk.text[i:i + 10]
            time.sleep(0.01)
            yield chatbot

google_key_component = gr.Textbox(
    label = "GOOGLE API KEY",
    type = "password",
    placeholder = "...",
    visible = GEMINI_API_KEY is None
)

image_prompt_component = gr.Image(
    type = "pil",
    label = "Image"
)
chatbot_component = gr.Chatbot(
    # label = 'Gemini',
    bubble_full_width = False
)
text_prompt_component = gr.Textbox(
    placeholder = "Chat with Gemini",
    label = "Ask me anything and press Enter"
)
run_button_component = gr.Button(
    "Run"
)
temperature_component = gr.Slider(
    minimum = 0,
    maximum = 1.0,
    value = 0.5,
    step = 0.05,
    label = "Temperature"
)
max_output_tokens_component = gr.Slider(
    minimum = 1,
    maximum = 8192,
    value = 4096,
    step = 1,
    label = "Max Output Tokens"
)
stop_sequences_component = gr.Textbox(
    label = "Add stop sequence",
    placeholder = "STOP, END"
)
top_k_component = gr.Slider(
    minimum = 1,
    maximum = 40,
    value = 32,
    step = 1,
    label = "Top-K"
)
top_p_component = gr.Slider(
    minimum = 0,
    maximum = 1,
    value = 1,
    step = 0.01,
    label = "Top-P"
)

user_inputs = [
    text_prompt_component,
    chatbot_component
]
bot_inputs = [
    google_key_component,
    image_prompt_component,
    temperature_component,
    max_output_tokens_component,
    stop_sequences_component,
    top_k_component,
    top_p_component,
    chatbot_component
]

with gr.Blocks() as demo:
    gr.HTML(TITLE)
    with gr.Row():
        gr.Markdown(NOTICE)
        gr.Markdown(ABOUT)
        gr.Markdown(ERRORS)
        gr.Markdown(FUTURE_IMPLEMENTATIONS)
    with gr.Column():
        google_key_component.render()
        with gr.Row():
            image_prompt_component.render()
            chatbot_component.render()
        text_prompt_component.render()
        run_button_component.render()
        with gr.Accordion("Parameters", open=False):
            temperature_component.render()
            max_output_tokens_component.render()
            stop_sequences_component.render()
            with gr.Accordion("Advanced", open=False):
                top_k_component.render()
                top_p_component.render()

    run_button_component.click(
        fn = user,
        inputs = user_inputs,
        outputs = [
            text_prompt_component,
            chatbot_component
        ],
        queue = False
    ).then(
        fn = bot,
        inputs = bot_inputs,
        outputs = [
            chatbot_component
        ]
    )
    text_prompt_component.submit(
        fn = user,
        inputs = user_inputs,
        outputs = [
            text_prompt_component,
            chatbot_component
        ],
        queue = False
    ).then(
        fn = bot,
        inputs = bot_inputs,
        outputs = [
            chatbot_component
        ]
    )

demo.launch()